/*

Margaret Brower	
Proposal Title: Reframing Women’s Issues: How Intersectional Identity Frames affect Women’s Political Attitudes	


HYPOTHESES

Stated-Hyp1: "The Policy Support Hypothesis, white women exposed to the “white woman/ white collar” policy frame will support gender-related policies more, relative to those exposed to the low-wage and nonwhite frames."

	Test-Hyp1: white women exposed to the “white woman/ white collar” policy frame will support gender-related policies more, relative to those exposed to the low-wage and nonwhite frames

Stated-Hyp2: "The Issue Prioritization Hypothesis, white women exposed to the “white woman/white collar” policy frame will prioritize gender-related issues over other political issues more, relative to those exposed to the low-wage and nonwhite frames."

	Test-Hyp2: white women exposed to the “white woman/white collar” policy frame will prioritize gender-related issues over other political issues more, relative to those exposed to the low-wage and nonwhite frames.

Stated-Hyp3: "The Policy Impact Hypothesis, white women exposed to the “white woman/white collar” policy frame will believe gender-related policies benefit all women regardless of class or race, whereas white women exposed to the low-wage and nonwhite frames will believe these policies benefit a sub-group population of women."

	// can't test. Author's proposed outcome variable (see proposal p. 10) was not fielded. 
	
	
Stated-Hyp4: "Difference of Issues Hypothesis, white women exposed to the low-wage and nonwhite frames will be less supportive of the wage gender-gap issue, relative to those who received the sexual harassment issue"
	
	Test-Hyp3: support for gender-related policies will be lower among white women exposed to the wage gender-gap issue (low-wage and non-white frames only), relative to those who received the sexual harassment issue
	

********************************************************************************
* NOTES
- sample is restricted to white women
- to see why/how outcome variables are selected for hypothesis 1, see page 9 of
 proposal	
*/

clear all
use "BrowerS27.dta", clear

********************************************************************************

* INDICATORS OF EXPERIMENTAL MANIPULATIONS

tab P_BROWER	
/*
Brower condition	
		1	SHC control
		2	SH1 white woman white collar
		3	SH2 minority woman white collar
		4	SH3 white woman pink/blue collar
		5	SH4 minority woman pink/blue collar
		6	GWGC control
		7	GWG1 white woman white collar
		8	GWG2 minority woman white collar
		9	GWG3 white woman pink/blue collar
		10	GWG4 minority woman pink/blue collar
		
SH : sexual harassment
GWG: gender wage gap		
see quex for details about what conditions mean
*/

* "white woman/ white collar" vs. "low-wage and nonwhite frames".
	lab def frame 1 "white OR white collar" 0 "non-white AND low wage"
	recode P_BROWER (2/3=1) (7/8=1) (5=0) (9=0) (*=.), gen(white_or_prof)
	lab val white_or_prof frame
	tab white_or_prof
	
* issueframe (hyp 4)
	lab def frame2 1 "Wage gap (low-wage AND non-white only)" 0 "sexual harassment"
	recode P_BROWER	(10=1) (1/5=0) (*=.), gen(issueframe)
	lab val issueframe frame2
	tab issueframe
	

* OUTCOME MEASURES

 
* outcome 1: support gender-related policies 

	/*Q1	To what extent would you oppose or support the proposed policy agenda to address [sexual harassment/the gender wage gap]?
	*/

	clonevar policysupport=Q1
	replace policysupport=. if Q1>7
	tab policysupport
	
* outcome 2: prioritize gender-related issues over other political issues more

		
	/*
	Q3.
	Imagine you are given a budget of $300.00 to allocate among the five organizations below. What amount of the budget would you allocate to each organization? You can only allocate to the below and you must allocate <u>ALL</u> of the budget. Please list what amount you would give to each. The total must sum to $300.00.

	- note: question is included in data but missing from codebook.
	*/

	* support for  women's orgs: option 2 and 5
	
	// recode missing
	foreach var of varlist Q3_* {
	replace `var'=. if `var'==998
	}
	
	gen fundwomen=Q3_2 +Q3_5
	
	* non-women orgs: 1, 3, 4
	gen fundother=Q3_1 +Q3_3 + Q3_4
	
	* measure of gender prioritization: difference of women and other
	gen genderpriority = fundwomen - fundother
	tab genderpriority

 
* outcome 3: believe gender-related policies benefit all women regardless of class or race 
/*
Q2A	[Women of color] If the proposed policy agenda by the organization [Women Against Sexual Violence & Harassment/Women for Equal Pay] is implemented, how will this policy impact the following groups …
	
Q2B	[White women] If the proposed policy agenda by the organization [Women Against Sexual Violence & Harassment/Women for Equal Pay] is implemented, how will this policy impact the following groups …
*/

	* recode missing
	replace Q2A=. if Q2A>5
	replace Q2B=. if Q2A>5
	
	* not sure what to do with this.

	
********************************************************************************

* ANALYSIS
	
*Test-Hyp1: white women exposed to the “white woman/ white collar” policy frame will support gender-related policies more, relative to those exposed to the low-wage and nonwhite frames
	
	reg policysupport i.white_or_prof
	// reject. p=0.104
	tess 1.white_or_prof +, init(BrowerS27) bonf(2)
	
*Test-Hyp2: white women exposed to the “white woman/white collar” policy frame will prioritize gender-related issues over other political issues more, relative to those exposed to the low-wage and nonwhite frames.

	reg genderpriority i.white_or_prof
	// reject. p=0.315
	tess 1.white_or_prof +, bonf(2)
	
*Test-Hyp3: support for gender-related policies will be lower among white women exposed to the wage gender-gap issue (low-wage and non-white frames only), relative to those who received the sexual harassment issue
	
	reg policysupport i.issueframe
	// reject. p=0.836
	tess 1.issueframe -
